Package org.vietspider.html.parser

Examples of org.vietspider.html.parser.NodeImpl

92
93
94
95
96
97
98
99
100
101
102
103
104
//    resources.add(new Resource("link", "href"));
//    resources.add(new Resource("script", "src"));
    downloadResources(address, tokens, resources);

    for(int i = 0; i < tokens.size(); i++) {
      NodeImpl token = tokens.get(i);
      if(token.getType() != TypeToken.TAG || !token.isNode(Name.A)) continue;
      Attributes attributes = token.getAttributes();
      Attribute attribute = attributes.get("href");
      if(attribute == null) continue;
      String link  = attribute.getValue();
      if(link == null || link.trim().length() < 1) continue;
      link  = urlUtils.createURL(parent, link);
View Full Code Here
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
    builder.append('>');
  }

  public void downloadResources(String referer, List<NodeImpl> tokens, List<Resource> resources) {
    for(int i = 0; i < tokens.size(); i++) {
      NodeImpl token = tokens.get(i);
      if(token.getType() != TypeToken.TAG) continue;
      for(Resource resource : resources) {
        if(!token.isNode(resource.tag))  continue;
        try {
          Attributes attributes = token.getAttributes();
          for(Attribute attribute : attributes) {
            if(!attribute.getName().equalsIgnoreCase(resource.attr))  continue;
            String rscName = toName(attribute.getValue());
            File file = new File(folder, rscName);
            String link = urlUtils.createURL(url, attribute.getValue());
View Full Code Here
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
  public List<Form> searchForm(List<NodeImpl>  tokens) {
    List<Form> forms = new ArrayList<Form>();
   
    Form form = null;
    for(int i = 0; i < tokens.size(); i++) {
      NodeImpl node = tokens.get(i);
     
      if(node.isNode(Name.FORM)) {
//        System.out.println("== > thay "+ node + " : "+ node.isOpen()  );
        if(node.getType() == TypeToken.TAG) {
          if(form != null) forms.add(form);
          form = new Form();
          Attributes attributes = node.getAttributes();
          Attribute attribute = attributes.get(NAME_ATTR);
          if(attribute != null) form.setName(attribute.getValue());
          attribute = attributes.get("method");
          if(attribute != null) form.setMethod(attribute.getValue());
          attribute = attributes.get(ACTION);
          if(attribute != null) form.setAction(attribute.getValue());
        } else {
          if(form != null) forms.add(form);
          form = null;
        }
      } else if(node.isNode(Name.INPUT)
          || node.isNode(Name.TEXTAREA)
          || node.isNode(Name.SELECT)) {       
        if(form == null) continue;
       
        Attributes attributes = node.getAttributes();
        Attribute attribute = attributes.get(NAME_ATTR);
        if(attribute == null) attribute = attributes.get(ID_ATTR);;
        if(attribute == null) continue;
        String name  = attribute.getValue();
        if(name == null || name.trim().isEmpty()) continue;
View Full Code Here
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
      message = "Error: Can't parse tokens!";
      return ERROR;
    }

    for(int i = 0; i < tokens.size(); i++) {
      NodeImpl node = tokens.get(i);
      if(!node.isNode(Name.INPUT)) continue;
      String value = getAttribute(node, TYPE_ATTR);
      if(value == null) continue;
      if(value.equalsIgnoreCase(PASSWORD)) return LOGIN;
    }

    int i = 0;
    for(;i < tokens.size(); i++) {
      NodeImpl node = tokens.get(i);
     
      if(node.isNode(Name.FORM)) {
        if(formName == null) break;       
        Attributes attrs = node.getAttributes();
        String name = getAttribute(attrs, NAME_ATTR);
        if(formName.equalsIgnoreCase(name)) break;
      }
    }

    HTMLNode form = null;
    List<HTMLNode> inputs = new ArrayList<HTMLNode>();

    for(; i < tokens.size(); i++) {
      NodeImpl node = tokens.get(i);
      if(node.isNode(Name.FORM)) {
        if(node.isOpen()) {
          form = node;
        } else {
          break;
        }
      } else if(node.isNode(Name.INPUT)
          || node.isNode(Name.TEXTAREA)
          || node.isNode(Name.SELECT)) {       
        inputs.add(node);
      }
    }
   
    if(form == null || inputs.size() < 1) {
View Full Code Here
328
329
330
331
332
333
334
335
336
337
338
339
    char [] chars = CharsDecoder.decode(charset, data, 0, data.length);
    List<NodeImpl> tokens  = parser.createTokens(chars);
    if(tokens == null) return true;

    for(int i = 0; i < tokens.size(); i++) {
      NodeImpl node = tokens.get(i);
      if(!node.isNode(Name.INPUT)) continue;
      String value = getAttribute(node, TYPE_ATTR);
      if(value == null) continue;
      if(value.equalsIgnoreCase(PASSWORD)) return true;
    }
    return false;
View Full Code Here
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
      message = "Can't parse tokens!";
      return ERROR;
    }

    for(int i = 0; i < tokens.size(); i++) {
      NodeImpl node = tokens.get(i);
      if(!node.isNode(Name.INPUT)) continue;
      String value = getAttribute(node, TYPE_ATTR);
      if(value == null) continue;
      if(value.equalsIgnoreCase(PASSWORD)) return LOGIN;
    }

    int i = 0;
    for(;i < tokens.size(); i++) {
      NodeImpl node = tokens.get(i);
      if(node.isNode(Name.FORM)) break;
    }

    HTMLNode form = null;
    List<HTMLNode> inputs = new ArrayList<HTMLNode>();

    for(; i < tokens.size(); i++) {
      NodeImpl node = tokens.get(i);
      if(node.isNode(Name.FORM)) {
        if(node.isOpen()) {
          form = node;
        } else {
          break;
        }
      } else if(node.isNode(Name.INPUT)
          || node.isNode(Name.TEXTAREA)
          || node.isNode(Name.SELECT)) {       
        inputs.add(node);
      }
    }
    if(form == null || inputs.size() < 1) {
      message = "form not found";
View Full Code Here
77
78
79
80
81
82
83
84
85
86
87
88
          if(type == RENDERER) builder.append('\n');
        }
        break;
      case SCRIPT:
      case STYLE:
        NodeImpl nodeImpl = (NodeImpl) node;
        if(nodeImpl.getType() == TypeToken.TAG && iterator.hasNext()) iterator.next();
        break;
      default:
        if(builder.length() > 0) {
          char c = builder.charAt(builder.length()-1);
          if(!(Character.isWhitespace(c)
View Full Code Here
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
    if(start == -1) {
      throw new UnknownHostException("Not found login form. Please check login address: "+loginUrl);
    }

    for(; start > -1; start--) {
      NodeImpl node = tokens.get(start);
      if(node.isNode(Name.FORM)) break;
    }

    HTMLNode form = null;
    boolean md5 = false;
    List<HTMLNode> inputs = new ArrayList<HTMLNode>();
    String formValue = null;

    for(int i = start; i < tokens.size(); i++) {
      NodeImpl node = tokens.get(i);
      if(node.isNode(Name.FORM)) {
        if(node.isOpen()) {
          if(!md5) {
            String value = new String(node.getValue());
            md5 = value.toLowerCase().indexOf("md5") > -1;
          }
          form = node;
          formValue = new String(form.getValue()).toLowerCase();
        } else {
          break;
        }
      } else if(node.isNode(Name.INPUT)) {
        if(!md5 && formValue != null) {
          md5 = formValue.indexOf("md5") > -1;
        }
        inputs.add(node);
      }
View Full Code Here
277
278
279
280
281
282
283
284
285
286
287
288
    return buffer.toString();
  }
 
  private int searchPasswordField(List<NodeImpl> tokens) {
    for(int i = 0; i < tokens.size(); i++) {
      NodeImpl node = tokens.get(i);
      if(!node.isNode(Name.INPUT)) continue;
      String value = getAttribute(node, TYPE_ATTR);
      if(value == null) continue;
      if(value.equalsIgnoreCase(PASSWORD)) return i;
    }
    return -1;
View Full Code Here
84
85
86
87
88
89
90
91
92
93
94
95
      case TD:
        separateBlock(builder, node, 2, wrappers);
        break;
      case SCRIPT:
      case STYLE:
        NodeImpl nodeImpl = (NodeImpl) node;
        if(nodeImpl.getType() == TypeToken.TAG && iterator.hasNext()) iterator.next();
        break;
      default:
        if(builder.length() > 0) {
          char c = builder.charAt(builder.length()-1);
          if(!(Character.isWhitespace(c)
View Full Code Here
TOP

Related Classes of org.vietspider.html.parser.NodeImpl

  • org.vietspider.browser.form.FormUtils
  • org.vietspider.browser.LoginWebsiteHandler
  • org.vietspider.browser.VirtualFormPost
  • org.vietspider.browser.VirtualFormPostBak
  • org.vietspider.chars.CharsEncoder
  • org.vietspider.html.NodeConfig
  • org.vietspider.html.renderer.ContentRendererBak
  • org.vietspider.html.renderer.TextRenderer
  • org.vietspider.html.util.HTMLAnchorUtil
  • org.vietspider.html.util.HyperLinkUtil
    • Copyright © 2018 www.massapicom. All rights reserved.
    All source code are property of their respective owners. Java is a trademark of Sun Microsystems, Inc and owned by ORACLE Inc. Contact coftware#gmail.com.